How and Why We Rebuilt Auto-scaling in OpenFaaS with Prometheus

Explore the journey of rebuilding the auto-scaling subsystem in OpenFaaS with Prometheus in this informative conference talk. Discover how customer feedback prompted the development of an improved version that better serves user needs, including the ability to scale based on in-flight requests and CPU usage, in addition to requests per second. Gain insights into the challenges faced during the process, learn from the PromQL queries developed, and understand the methods used for data collection and instrumentation. Witness live demonstrations of function scaling tied to real-world use cases, and acquire valuable knowledge about enhancing serverless architectures for better performance and efficiency.

How and Why We Rebuilt Auto-scaling in OpenFaaS with Prometheus - Alex Ellis, OpenFaaS Ltd